Maternal Health Task Force


How Researchers Used Behavioral Economics to Help Women Decide Where to Give Birth in Nairobi, Kenya

By: Sarah Hodin, Project Coordinator II, Women and Health Initiative, Harvard T.H. Chan School of Public Health

Jessica Cohen, Katherine Lofgren and Margaret McConnell recently published a paper in the American Economic Review titled, “Precommitment, cash transfers, and timely arrival for birth: Evidence from a randomized controlled trial in Nairobi Kenya.” The Maternal Health Task Force (MHTF)’s Sarah Hodin interviewed Jessica Cohen and Margaret McConnell about their study.

SH: What was the impetus for this study?

JC: The idea for this study started when we were working in peri-urban Nairobi, Kenya and found that very late in their pregnancies, women were still considering giving birth at several different facilities. In this area, there are roughly 800 facilities that offer labor and delivery care, so women have an overwhelming number of choices. We were surprised to see how late most women were making decisions about delivery location and how few ended up going where they thought they would or where they themselves wanted to deliver during pregnancy. These last-minute decisions and changes in birth plans could lead to delays in arrival and potentially to women choosing poorer quality facilities. The aim of our study was to test an intervention based on behavioral economics principles to make the birth planning process more thorough and deliberate. Our ultimate goal was to help women deliver at higher quality facilities that were more in line with their preferences during pregnancy and to hopefully encourage earlier arrival at the delivery facility.

SH: Tell me about the study design and methods you used.

JC: This was a proof-of-concept study partially funded through a Grand Challenges Exploration grant from the Bill & Melinda Gates Foundation. We recruited women from 24 different neighborhoods in Nairobi by consulting community health workers’ lists of pregnant women, organizing recruitment events and using snowball sampling. Women who chose to participate in the study were randomized into three groups, but we’ll talk about results from the two groups that were included in the American Economic Review paper: a control arm and a treatment arm.

We interviewed all of the women when they were recruited at baseline, during the eighth month of pregnancy and 2-4 weeks postpartum. During the eight-month visit, women in the treatment group were asked to pre-commit to a delivery location and prepare a couple of back-up options. Those women were given a non-conditional, labelled cash transfer that we suggested they use to deliver at the place they wanted. Women who ended up giving birth at the facility where they pre-committed were given an additional cash transfer following delivery.

MM: It may be important to note that only women who wanted to deliver in a facility were included in this study. So, not surprisingly, only a handful of the women who participated ended up giving birth at home.

SH: Are these commonly-used techniques in behavioral economics?

JC: Yes, the pre-commitment and cash transfer components are often used as strategies to influence people’s health-seeking behaviors.

MM: In the context of our study, the idea behind this intervention was to encourage women to make decisions about delivery location earlier, to choose places in line with their own preferences and to follow through with those decisions.

SH: What were your main findings?

JC: The American Economic Review paper focused on a subset of our findings that were related to delays. While delays are discussed quite a bit in the maternal health literature, we didn’t find any papers where delays were the outcome of the analysis, and very few papers even quantified the extent of delays. At what point does a woman’s arrival at a facility during labor constitute a delay? There is very little technical guidance on this.

MM: The question we couldn’t answer based on the existing literature was, what measure should be used to determine if a delay occurred? We talked to a lot of obstetricians about this and figured out what data we could feasibly collect from interviews with mothers in order to operationalize and measure delays.

JC: We came up with an index that combines several aspects of delays: time between contractions when departing home for the facility and upon arrival, cervical dilation at first exam and time between arrival at the facility and delivery. The treatment group had significantly fewer delays overall based on this index. The average contraction spacing at departure and upon arrival were significantly longer among women in the treatment group, which is an indication of a shorter delay. Also, significantly more women in the treatment group delivered in the facility that they most wanted or at a facility that were considering during pregnancy.

SH: What are the implications of this study? Do you think these findings are unique to Nairobi?

MM: The findings are certainly relevant in contexts where women have numerous choices, which is the case in many urban settings.

JC: There is definitely a need for interventions that relieve financial barriers to facility-based delivery, but there are also non-monetary strategies we can use to help improve outcomes. Pre-commitment is one of these. However, the problem remains that we don’t really know how late is too late in terms of implications for maternal and newborn health outcomes. Agreement on a concrete and operational definition of life-threatening “delays” would help researchers design studies that consider approaches to reducing delays and thereby improve delivery outcomes.

Other results from this work are forthcoming.

Access this paper by Jessica Cohen, Katherine Lofgren and Margaret McConnell.

Learn more about demand-side financing strategies for improving maternal health.

Subscribe to receive new posts from the MHTF blog in your inbox.

Categories: Maternal Health


Post navigation